- Email: [email protected]
On the fidelity of the lactoperoxidase method of cell membrane radioiodination: an electron microscopic autoradiographic study
Journal of Immunological Methods, 59 (1983) 1- 11
1
Elsevier Biomedical Press
On the Fidelity of the Lactoperoxidase Method of Cell Membrane Radioiodination: an Electron Microscopic Autoradiographic Study W. Cushley l,,, J.R.J. Baker 2,**, I.F. Hassan 2 and I.H.M. Williamson 2 I Department of Biochemistry, University of Glasgow, Glasgow G12 8QQ, and 2 Ciba-Geigy Pharmaceuticals Division, Research Centre, Wimblehurst Road, Horsham RH12 4,41t, U.K.
(Received 19 May 1982, accepted 20 November 1982)
The radiolabelling of human peripheral blood lymphocytes by lactoperoxidase-catalysed iodination using 2 different sources of hydrogen peroxide has been compared using electron microscopic autoradiography. A method of statistical analysis of the autoradiographs has permitted precise identification of radioactive sources, in particular cellular compartments, taking into account cross-fire of Auger electrons producing silver grains in compartments other than those from which they are emitted. Our data confirm the postulates of previous investigators that the majority of radioiodine is located at the plasma membrane of cells labelled by enzymic iodination. The results further suggest that the glucose-glucose oxidase system for generation of hydrogen peroxide permits a greater degree of specific radiolabeUingof plasma membrane proteins with less damage than equivalent lactoperoxidase iodination reactions promoted by exogenously added hydrogen peroxide. Key words: lactoperoxidase labelling - - membrane radiolabelling - - electron microscopy autoradiography
Introduction The p l a s m a m e m b r a n e is the cell's m e a n s of i m m e d i a t e c o m m u n i c a t i o n with its m i c r o e n v i r o n m e n t . C o n s e q u e n t l y , the macromolecules f o u n d o n the cell surface are of i m p o r t a n c e in cellular c o m m u n i c a t i o n processes, such as transport of metabolites, r e c o g n i t i o n of foreign material a n d receptors for hormones, drugs a n d viruses. Study of the structure a n d f u n c t i o n of such cell surface macromolecules is often performed b y trace radiolabelling in situ of the m a c r o m o l e c u l e of interest. Clearly, it is desirable to study these macromolecules as they are expressed o n the cell surface a n d so radiolabelling procedures which do n o t label cytoplasmic macromolecules are
* Present address: Department of Microbiology, University of Texas Health Science Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas 75235, U.S.A. **To whom correspondence should be addressed. 0022-1759/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
necessary. Several such procedures have been described. The use of lactoperoxidase to catalyse radioiodination of immunoglobulins and other proteins in vitro was first reported by Marchalonis (1969). The lactoperoxidase-catalysed radioiodination technique (Marchalonis et al., 1971) does label cell membrane proteins, but criticisms have been made regarding the amounts of exogenously added hydrogen peroxide used which might damage the cell membrane sufficiently to allow access of isotope to the cytoplasm. Provision of hydrogen peroxide by a more controlled means has been achieved by using a glucose plus glucose oxidase coupled reaction (G-GO) (Hubbard and Cohn, 1975). Since this procedure generates hydrogen peroxide in minimal, but sufficient quantities for the radiolabelling reaction to proceed, it may lead to less damage of the cell membrane and, consequently, less radiolabelling of cytoplasmic components. That cell viability was not significantly affected by the source of hydrogen peroxide (Gonatas et al., 1976) was illustrated by the similar capability of rat splenocytes to exclude trypan blue irrespective of the iodination regime. The purpose of the present study was to determine the relative extent of cell membrane and intracellular protein in human peripheral lymphoc~ctes and a /3lymphoma cell line labelling. This was achieved by lactoperoxidase-catalysed iodination employing hydrogen peroxide supplied exogenously or generated by the G-GO reaction. The technique chosen was electron microscopic (EM) autoradiography using a method of analysis (Blackett and Parry, 1973, 1977) designed to account totally for cross-fire from sources of regular or irregular geometry. This approach confirms and extends the results of Gonatas et al. (1976) who used the grain distribution about a linear source of 125I to determine the putative cellular location of the radioactivity.
Materials and Methods
Cell culture The B-lymphoma cell line Daudi was maintained as a suspension culture in RPMI 1640 medium supplemented with 10% (v/v) foetal calf serum and 2 mM fresh glutamine. Penicillin and streptomycin were included at final concentrations of 100 U / m l and 100/~g/ml respectively (complete medium). The cultures were kept in a 95% relative humidity incubator at 37°C with a 5% CO 2 atmosphere.
Preparation of peripheral blood lymphocytes (PBLs) Blood was obtained from normal healthy volunteers by venipuncture, anticoagulated in EDTA (0.2%) and diluted with an equal volume of RPMI 1640 medium supplemented as above without foetal calf serum. The cell suspension was then carefully layered on Ficoll-Hypaque (Ficoll-Paque ®, Pharmacia) and the mixtures centrifuged at 250 x g for 30 min (Bryum, 1968). The lymphocyte-rich mononuclear cells were harvested from the interface. The lymphocytes were washed twice with complete medium and incubated at 37°C for 30 rain in a plastic tissue culture flask to allow removal of adherent cells. The non-adherent cells were aspirated and prepared for radioiodination as described below.
Radioiodination procedures For radioiodination, 107 cells from log-phase cultures of Daudi cells, o r 5 - ] 0 6 (PBLs) from Ficoll-Hypaque gradients were used. In each case the viability of the cells was assessed by trypan blue dye exclusion and only cultures of > 95% viability were employed for radioiodination. The cells were washed twice with 25 ml aliquots of serum-free supplemented RPMI 1640 and once with 25 ml of phosphate-buffered saline (PBS) prior to transfer to a 1.5 ml capacity polypropylene tube. The cells were resuspended in lactoperoxidase (50/~1 of a 1 m g / m l solution) in PBS. Carrier-free Na125I (100 ~Ci) was then added. (a) Reaction with added 1120e. Radioiodination was initiated by addition of 10 /~l of a freshly prepared 1:1000 dilution of H202 (100 vols.) in PBS. The reaction was allowed to proceed at room temperature for 3-4 min and radiolabelling terminated by addition of 5 mM potassium iodide in PBS 1 ml. After standing for 2 min to displace free radioiodide, the labelled cells were washed 3 times and finally suspended in 1 ml of complete medium prior to fixation. (b) Reaction with glucose-glucose oxidase. Glucose oxidase in PBS (10 #1 of a 1 m g / m l solution) was added to the cells and the reaction started by addition of glucose (10 t~l of 50 mM) in PBS. The radiolabelling was allowed to proceed for 20 rain at room temperature with occasional mixing of the contents of the tube and was terminated as described above. (e) Control experiment. PBLs were incubated with Na125I alone, omitting both lactoperoxidase and any hydrogen peroxide.
Preparation of cells for electron microscopy Daudi cells in PBS (1 ml) or human PBLs in 1 ml of complete medium were added to 2,5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2 (2 ml) and left at room temperature for 5 min. The samples were then centrifuged at 90 × g for 15 rain. The supernatant was discarded and the pellet resuspended in a minimum volume (50-100/xl) of bovine serum albumin (10% w / v in sterile saline solution). This suspension was rapidly gelled with 12.5% glutaraldehyde in cacodylate buffer (Bullock and Christian, 1976). The gelled pellet was post-fixed for 30 rain in a 1 : 2 ( v / v ) mixture of glutaraldehyde (2.5%) and osmium tetroxide (1%) in cacodylate buffer. Following fixation, the specimens were rinsed twice for 15 min each in cacodylate buffer with a third rinse overnight. The platelets were dehydrated in a graded ethanol series (Table I), treated with propylene oxide for 5 min and infiltrated with epon/araldite (Mollenhauer, 1964) before final embedding and polymerization.
Determination of recovery of radioactivity from specimens during preparation for electron microscopy The procedures described above were repeated in triplicate in 5 ml aliquots of the processing solutions. In order to count the radioactivity in the final pellet, this was digested in 1 N sodium hydroxide. Samples from each stage of the processing schedule including the final pellets were counted for 1 min in a Beckman biogamma
TABLE I RECOVERY OF RADIOACTIVITY D U R I N G PREPARATION F O R ELECTRON MICROSCOPIC A U T O R A D I O G R A P H Y OF H U M A N PERIPHERAL LYMPHOCYTES LABELLED BY THE LACTOPEROXIDASE M E T H O D Processing solution
% Total recovered radioactivity (expressed as a mean; n = 3)
Glutaraldehyde Glutaraldehyde + osmium tetroxide Cacodylate 1 2 3 70% Ethanol 90% Ethanol Abs Ethanol 1 Abs Ethanol 2 Propylene oxide Pellet
68.4 11.3 1.3 0.3 1.0 1.3 0.6 0.2 0.1 0.3 15.3
counter and the mean recovery of radioactivity at each stage was expressed as a percentage of the total (Table I).
Preparation of specimens for electron microscopic autoradiography The method used to prepare EM autoradiographs has been described previously (Baker et al., 1977). In the present study, exposure was from 1-4 weeks. The specimens ultimately used for analysis were developed in Kodak D 19. Cells associated with a minimum of 8 silver grains were photographed at random. The final print magnification was X 25,000.
Analysis of EM autoradiographs It is now well established that a majority of silver grains in EM autoradiographs do not directly overlie the point in the tissue section from which the isotopic decay particles responsible for their formation arose. This is the result of a number of factors, such as variable electron energy, section and emulsion thickness, angle of emission and above all, the ability of the electron microscope to resolve this 'range distribution' of silver grains. A number of approaches have been proposed which are designed to account for this 'cross-fire' from a source of decay which may be in organelle a, to the site of the silver grain which may be over organelle b. In the present study we have used the hypothetical grain methods of Blackett and Parry (1973, 1977). In this method, a computer programme is used to compare hypothetical grain distributions with the real grain distribution. According to the hypothesis being tested, hypothetical sources of radioactivity are established by applying to the autoradiographs a transparent overlay screen which contains computer-predicted source-to-site dis-
tances which generate the hypothetical silver grains 'emanating' from these sources. The directions of these hypothetical decays are, of course, random. The distances are derived from experimentally measured range distribution curves for line sources of a given isotope of the type first reported by Salpeter et al. (1969). Hence a rectangular matrix of cross-fire is created, there being invariably more sites than sources. Estimated of radioactivity in the various sources derive from systematic modification of the hypothetical source values by a minimising sub-routine until the hypothetical grain and real grain distribvtions fit optimally for the designated sites as assessed by the chi-square test. A feature of this method is that any hypothesis of radiolabel distribution can be tested. In cases where the crude location of silver grains is difficult to interpret by visual inspection, it is usual to collect hypothetical grain data by random or regular sampling so as to include all the potential sources of radioactivity in the analysis. In this study, however, many of the silver grains were sited close to the plasma membrane. Therefore, in order to collect sufficient relevant hypothetical grains for this source, which occupies a negligible area, regular points on the membrane were sampled for cross-fire (hypothetical grains) in addition to regular sampling of the rest of the cell for other potential sources. It should be stated that all of the real grains must be, and were, counted. This type of 'weighted' hypothetical grain sampling means that the relative area of each potential source is not realistically represented. Hence, only the computer estimates expressed as 'relative activity' (% of total, see Blackett and Parry, 1977) are valid. For the 4 experiments reported, the numbers of hypothetical and real grains counted ranged from 1161-1311 and 400-615 respectively.
Results
Retention of radioactivity by the EM preparative procedure The elution of counts from the radioiodinated cell suspension during fixation and dehydration was monitored by direct gamma counting of triplicate samples treated TABLE II HYPOTHETICAL GRAIN ANALYSIS OF LACTOPEROXIDASE A N D EXOGENOUS H202 Source
% Total radioactivity
Plasma membrane Cytosol Mitochondria Nucleus Matrix
54.7 + 4.7 17.3 + 2.8 1.8 ___1.1 7.7 + 1.4 18.5 ± 3.9
Total X2 1.9 (3 DF)
P > 0.5
125I-LABELLED
DAUDI
CELLS
USING
6 TABLE III HYPOTHETICAL GRAIN ANALYSIS OF 125I-LABELLED HUMAN LACTOPEROXIDASE AND EXOGENOUS HzOz Source
% Total radioactivity
Plasma membrane Cytosol Nucleus Matrix
67.0 _+4.5 16.2 + 5.5 4.3 _+2.5 12.6 + 1.1
Total X~ 1.3 (1 DF)
P > 0.20
PBLs USING
identically to those prepared for electron microscopy (Table I). A l t h o u g h 80% of radiolabel appeared to be lost i n the 2 fixation stages (glutaraldehyde; glutaraldeh y d e + o s m i u m tetroxide) this was unlikely to be a t t r i b u t a b l e to loss of p r o t e i n - b o u n d radioiodine. Rather, it was more likely to be the result of loss of radioactivity from the surface connected system of c o n t a m i n a t i n g platelets a n d radioiodine, possibly associated with lipid. The cacodylate rinses a n d ethanol d e h y d r a t i o n remove 5% of total radiolabel.
Fig. 1. EM autoradiograph of human peripheral blood lymphocytes incubated with Na~:51 in the presence of the exogenouslyadministered H202 reaction. Membrane disruption is obvious and may contribute to the grain value for the matrix. × 19,100. n = nucleus, m-mitochondria.
7 TABLE IV HYPOTHETICAL GRAIN ANALYSIS OF J251-LABELLED HUMAN PBLs USING G-GO TO GENERATE H202 Source
% Total radioactivity
Plasma membrane Cytosol Nucleus
92.5 +__1.8 2.7 _+ 1.0 4.9 ± 0.7
Total X2 = 5.3 (4 DF)
P > 0.20
(a) Subcellular location of radioiodide following lactoperoxidase/exogenous I-t,0, catalysed-labelling Daudi cells. T h e results o f the h y p o t h e t i c a l g r a i n analysis for the c u l t u r e d cell line are s h o w n in T a b l e II. O n l y 55% o f t h e t o t a l c e l l - a s s o c i a t e d r a d i o l a b e l was a t t r i b u t a b l e to the cell m e m b r a n e w i t h m o s t o f the r e m a i n d e r d i s t r i b u t e d e q u a l l y b e t w e e e n t h e c y t o s o l a n d ' m a t r i x ' . W h i l s t t h e n u c l e u s c o n t a i n e d a l m o s t 8% the m i t o c h o n d r i a a c c o u n t e d for less t h a n 2% o f the total. T h e e x c e l l e n c e o f t h e fit
Fig. 2. EM autoradiograph of human peripheral blood lymphocytes incubated with Na 1251 in the presence of the G-GO reaction. The cell and cell membrane appears well fixed and almost all the silver grains are associated with the cell membrane. × 20,000.
8
TABLE V HYPOTHETICAL GRAIN ANALYSIS OF 1251-LABELLED HUMAN PBLs USING G-GO TO GENERATE H 202 Source
% Total radioactivity
Plasma membrane Cytosol Nucleus
89.0 + 0.8 8.9 +_0.5 2.1 +_0.4
total X2 = 2.0 (2 DF)
P > 0.20
b e t w e e n real a n d final hypothetical grains ( P > 0.5) suggests that the ' m a t r i x ' is genuinely labelled. However, it seems p r o b a b l e that m u c h of this c o m p o n e n t represents i n c o m p l e t e l y sloughed-off m e m b r a n e a n d / o r 'glycocalyx' debris produced as a consequence of the labelling technique. Peripheral blood lymphocytes (PBLs). T a b l e III shows the results for the equivalent r a d i o i o d i n a t i o n a n d analysis of PBLs. I n this i n s t a n c e 67% of the radiolabel was located in the cell m e m b r a n e . The values for cytosol (16.2%), nucleus (4.3%) a n d matrix (12.6%) are of a similar m a g n i t u d e to those in the D a u d i cells (see T a b l e II).
Fig. 3. EM autoradiograph of human peripheral blood lymphocytes incubated with Na~25I alone. Although the cells are well fixed they are devoid of silver grains. × 10,000.
Once again, there is fine structural evidence to demonstrate that a certain level of cell membrane disruption may substantially contribute to the value for the matrix (Fig. 1).
(b) Subcellular location of radioiodide following lactoperoxidase/glucose-glucose oxidase labelling The estimates of subcellular radioactivity for PBLs radioiodinated with lactoperoxidase and the G-GO coupled reaction for generation of H202 are expressed for duplicate experiments (Tables IV and V). Since the initial computer results revealed that matrix and mitochondria were unlabelled, these sources were omitted from the repeat analyses shown in order to provide a model of subcellular labelling which would give the best fit as judged by the chi-square test. In both experiments approximately 90% of radioactivity was associated with the plasma membrane of the PBLs with the remainder in the cytosol or nucleus. The hypothetical grain analysis revealed that the matrix was not labelled. Moreover, the infrequency of fine structural membrane damage in these experiments (Fig. 2) supports our observation that only where such damage occurs will the matrix appear to be labelled.
(c) Control experiment When PBLs were incubated with Na125I alone, the resultant autoradiographs prepared in an identical manner to those described above were largely devoid of silver grains (Fig. 3). The few silver grains which were observed appeared to be non-specifically distributed.
Discussion
The study of membrane proteins of small defined cell populations requires radiolabelling of these macromolecules to facilitate analyses. A prerequisite of such radiolabelling techniques including lactoperoxidase-catalysed radioiodination is that exclusively plasma membrane proteins should be labelled. We have therefore investigated 2 radioiodination techniques to assess the fidelity with which this goal may be achieved. Previous EM autoradiographic studies on the distribution of radioiodine in cellular compartments of cells labelled by the lactoperoxidase method have shown that the grains are not exclusively located at the cell membrane in spite of claims made to the contrary (Marchalonis et al., 1971; Hubbard and Cohn, 1975). Indeed, measured range distribution data for 1251 (Fertuck and Salpeter, 1976; Blackett et al., 1980) show that depending upon the autoradiographic conditions and, in particular, the emulsion type, the spread of all the grains from a known linear source is several times the 'half-distance' value (Salpeter et al., 1969). A subsequent study (Gonatas et al., 1976) employed histograms of grain density, measured from the plasma membrane, to predict whether or not the image spread was consistent with labelling exclusively in the plasma membrane. They concluded that since the
l0 intracellular grain distribution extended beyond that predicted by the line source data of Fertuck and Salpeter (1976), a proportion of the radioactive decays (36%) must be intracellular in origin. It is, however, questionable whether the convoluted plasma membrane of lymphocytes can be legitimately considered as a linear source. Consequently, it was our objective to employ a method of analysis which could assess the effect of cross-fire from sources of irregular geometry. We believe that the use of the hypothetical grain method enables a more precise estimate of activity to be determined in all the significantly radiolabelled sources. The use of a albumin/glutaraldehyde gel (Bullock and Christian, 1976) as a support medium for EM processing of cell suspensions not only serves to eliminate the need for potentially damaging repeated centrifugation steps, but also provides a means of spacing the cells such that cross-fire from one to another can be easily minimised. Analysis of cells radiolabeUed by lactoperoxidase catalysed iodination, the hydrogen peroxide being supplied either exogenously or from the G-GO reaction, demonstrated that in both cases, the highest percentage of the radioisotope was localised in the plasma membrane. A greater proportion of plasma membrane-associated radioactivity was demonstrable in cells where the hydrogen peroxide was derived from the G - G O reaction (Tables IV and V) than in those cells where the radiolabelling reaction was promoted by exogenously added hydrogen peroxide (Tables II and lID; clearly less cytoplasmic radioactivity was evident in the former cell population than in the latter. We conclude from these data that the G-GO system for generation of hydrogen peroxide in the lactoperoxidase catalysed iodination provides a method of labelling plasma-membrane proteins of greater fidelity than the equivalent reaction driven by exogenously supplied hydrogen peroxide. A possible explanation for the above observations is that the relatively high concentration of hydrogen peroxide in systems employing exogenously added catalyst may alter the permeability of the plasma-membrane to small ions without adversely affecting cell viability in the short term. In coupled systems where hydrogen peroxide is generated biochemically, the latter is probably utilised immediately by lactoperoxidase, thus precluding generation of high local concentrations of a potentially membrane-oxidising agent. In other words, it does not seem valid to assume a close correlation between cell viability, as assessed by dye exclusion tests (Gonatas et al., 1976), and the degree of intracellular penetration of radioiodine.
References
Baker, J.R.J., G.R. Bullock,N. Crawford and D.G. Taylor, 1977, Am. J. Pathol., 88, 277. Blackett, N.M. and D.M. Parry, 1973, J. Cell Biol. 57, 9. Blackett, N.M. and D.M. Parry, 1977, J. Histochem. Cytochern.25, 206. Blackett, N.M., D.M, Parry and J.R.J. Baker, 1980, J. Histochem. Cytochem.28, 1050. BiSyum,A., 1968, Seand. J. Clin. Lab. Invest. 21 (Suppl. 97), 77. Bullock, G.R. and R.A. Christian, 1976, Histochem. J. 8, 291.
11 Fertuck, H.C. and M.M. Salpeter, 1976, J. Cell. Biol. 69, 144. Gonatas, N.K., J.O. Gonatas, A. Stieber, J.C. Antoine and S. Avrameas, 1976, J. Cell. Biol. 70, 477. Hubbard, A.L. and Z.A. Cohn, 1975, J. Cell. Biol. 64, 438. Marchalonis, J.J., 1969, Biochem. J. 113, 299. Marchalonis, J.J., R.E. Cone and V. Santer, 1971, Biochem. J. 124, 921. Mollenhauer, H.H., 1964, Stain Technol. 39, 111. Salpeter, M.M., L. Bachmann and E.E. Salpeter, 1969, J. Cell Biol. 41, 1.
1
Elsevier Biomedical Press
On the Fidelity of the Lactoperoxidase Method of Cell Membrane Radioiodination: an Electron Microscopic Autoradiographic Study W. Cushley l,,, J.R.J. Baker 2,**, I.F. Hassan 2 and I.H.M. Williamson 2 I Department of Biochemistry, University of Glasgow, Glasgow G12 8QQ, and 2 Ciba-Geigy Pharmaceuticals Division, Research Centre, Wimblehurst Road, Horsham RH12 4,41t, U.K.
(Received 19 May 1982, accepted 20 November 1982)
The radiolabelling of human peripheral blood lymphocytes by lactoperoxidase-catalysed iodination using 2 different sources of hydrogen peroxide has been compared using electron microscopic autoradiography. A method of statistical analysis of the autoradiographs has permitted precise identification of radioactive sources, in particular cellular compartments, taking into account cross-fire of Auger electrons producing silver grains in compartments other than those from which they are emitted. Our data confirm the postulates of previous investigators that the majority of radioiodine is located at the plasma membrane of cells labelled by enzymic iodination. The results further suggest that the glucose-glucose oxidase system for generation of hydrogen peroxide permits a greater degree of specific radiolabeUingof plasma membrane proteins with less damage than equivalent lactoperoxidase iodination reactions promoted by exogenously added hydrogen peroxide. Key words: lactoperoxidase labelling - - membrane radiolabelling - - electron microscopy autoradiography
Introduction The p l a s m a m e m b r a n e is the cell's m e a n s of i m m e d i a t e c o m m u n i c a t i o n with its m i c r o e n v i r o n m e n t . C o n s e q u e n t l y , the macromolecules f o u n d o n the cell surface are of i m p o r t a n c e in cellular c o m m u n i c a t i o n processes, such as transport of metabolites, r e c o g n i t i o n of foreign material a n d receptors for hormones, drugs a n d viruses. Study of the structure a n d f u n c t i o n of such cell surface macromolecules is often performed b y trace radiolabelling in situ of the m a c r o m o l e c u l e of interest. Clearly, it is desirable to study these macromolecules as they are expressed o n the cell surface a n d so radiolabelling procedures which do n o t label cytoplasmic macromolecules are
* Present address: Department of Microbiology, University of Texas Health Science Center at Dallas, 5323 Harry Hines Boulevard, Dallas, Texas 75235, U.S.A. **To whom correspondence should be addressed. 0022-1759/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
necessary. Several such procedures have been described. The use of lactoperoxidase to catalyse radioiodination of immunoglobulins and other proteins in vitro was first reported by Marchalonis (1969). The lactoperoxidase-catalysed radioiodination technique (Marchalonis et al., 1971) does label cell membrane proteins, but criticisms have been made regarding the amounts of exogenously added hydrogen peroxide used which might damage the cell membrane sufficiently to allow access of isotope to the cytoplasm. Provision of hydrogen peroxide by a more controlled means has been achieved by using a glucose plus glucose oxidase coupled reaction (G-GO) (Hubbard and Cohn, 1975). Since this procedure generates hydrogen peroxide in minimal, but sufficient quantities for the radiolabelling reaction to proceed, it may lead to less damage of the cell membrane and, consequently, less radiolabelling of cytoplasmic components. That cell viability was not significantly affected by the source of hydrogen peroxide (Gonatas et al., 1976) was illustrated by the similar capability of rat splenocytes to exclude trypan blue irrespective of the iodination regime. The purpose of the present study was to determine the relative extent of cell membrane and intracellular protein in human peripheral lymphoc~ctes and a /3lymphoma cell line labelling. This was achieved by lactoperoxidase-catalysed iodination employing hydrogen peroxide supplied exogenously or generated by the G-GO reaction. The technique chosen was electron microscopic (EM) autoradiography using a method of analysis (Blackett and Parry, 1973, 1977) designed to account totally for cross-fire from sources of regular or irregular geometry. This approach confirms and extends the results of Gonatas et al. (1976) who used the grain distribution about a linear source of 125I to determine the putative cellular location of the radioactivity.
Materials and Methods
Cell culture The B-lymphoma cell line Daudi was maintained as a suspension culture in RPMI 1640 medium supplemented with 10% (v/v) foetal calf serum and 2 mM fresh glutamine. Penicillin and streptomycin were included at final concentrations of 100 U / m l and 100/~g/ml respectively (complete medium). The cultures were kept in a 95% relative humidity incubator at 37°C with a 5% CO 2 atmosphere.
Preparation of peripheral blood lymphocytes (PBLs) Blood was obtained from normal healthy volunteers by venipuncture, anticoagulated in EDTA (0.2%) and diluted with an equal volume of RPMI 1640 medium supplemented as above without foetal calf serum. The cell suspension was then carefully layered on Ficoll-Hypaque (Ficoll-Paque ®, Pharmacia) and the mixtures centrifuged at 250 x g for 30 min (Bryum, 1968). The lymphocyte-rich mononuclear cells were harvested from the interface. The lymphocytes were washed twice with complete medium and incubated at 37°C for 30 rain in a plastic tissue culture flask to allow removal of adherent cells. The non-adherent cells were aspirated and prepared for radioiodination as described below.
Radioiodination procedures For radioiodination, 107 cells from log-phase cultures of Daudi cells, o r 5 - ] 0 6 (PBLs) from Ficoll-Hypaque gradients were used. In each case the viability of the cells was assessed by trypan blue dye exclusion and only cultures of > 95% viability were employed for radioiodination. The cells were washed twice with 25 ml aliquots of serum-free supplemented RPMI 1640 and once with 25 ml of phosphate-buffered saline (PBS) prior to transfer to a 1.5 ml capacity polypropylene tube. The cells were resuspended in lactoperoxidase (50/~1 of a 1 m g / m l solution) in PBS. Carrier-free Na125I (100 ~Ci) was then added. (a) Reaction with added 1120e. Radioiodination was initiated by addition of 10 /~l of a freshly prepared 1:1000 dilution of H202 (100 vols.) in PBS. The reaction was allowed to proceed at room temperature for 3-4 min and radiolabelling terminated by addition of 5 mM potassium iodide in PBS 1 ml. After standing for 2 min to displace free radioiodide, the labelled cells were washed 3 times and finally suspended in 1 ml of complete medium prior to fixation. (b) Reaction with glucose-glucose oxidase. Glucose oxidase in PBS (10 #1 of a 1 m g / m l solution) was added to the cells and the reaction started by addition of glucose (10 t~l of 50 mM) in PBS. The radiolabelling was allowed to proceed for 20 rain at room temperature with occasional mixing of the contents of the tube and was terminated as described above. (e) Control experiment. PBLs were incubated with Na125I alone, omitting both lactoperoxidase and any hydrogen peroxide.
Preparation of cells for electron microscopy Daudi cells in PBS (1 ml) or human PBLs in 1 ml of complete medium were added to 2,5% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.2 (2 ml) and left at room temperature for 5 min. The samples were then centrifuged at 90 × g for 15 rain. The supernatant was discarded and the pellet resuspended in a minimum volume (50-100/xl) of bovine serum albumin (10% w / v in sterile saline solution). This suspension was rapidly gelled with 12.5% glutaraldehyde in cacodylate buffer (Bullock and Christian, 1976). The gelled pellet was post-fixed for 30 rain in a 1 : 2 ( v / v ) mixture of glutaraldehyde (2.5%) and osmium tetroxide (1%) in cacodylate buffer. Following fixation, the specimens were rinsed twice for 15 min each in cacodylate buffer with a third rinse overnight. The platelets were dehydrated in a graded ethanol series (Table I), treated with propylene oxide for 5 min and infiltrated with epon/araldite (Mollenhauer, 1964) before final embedding and polymerization.
Determination of recovery of radioactivity from specimens during preparation for electron microscopy The procedures described above were repeated in triplicate in 5 ml aliquots of the processing solutions. In order to count the radioactivity in the final pellet, this was digested in 1 N sodium hydroxide. Samples from each stage of the processing schedule including the final pellets were counted for 1 min in a Beckman biogamma
TABLE I RECOVERY OF RADIOACTIVITY D U R I N G PREPARATION F O R ELECTRON MICROSCOPIC A U T O R A D I O G R A P H Y OF H U M A N PERIPHERAL LYMPHOCYTES LABELLED BY THE LACTOPEROXIDASE M E T H O D Processing solution
% Total recovered radioactivity (expressed as a mean; n = 3)
Glutaraldehyde Glutaraldehyde + osmium tetroxide Cacodylate 1 2 3 70% Ethanol 90% Ethanol Abs Ethanol 1 Abs Ethanol 2 Propylene oxide Pellet
68.4 11.3 1.3 0.3 1.0 1.3 0.6 0.2 0.1 0.3 15.3
counter and the mean recovery of radioactivity at each stage was expressed as a percentage of the total (Table I).
Preparation of specimens for electron microscopic autoradiography The method used to prepare EM autoradiographs has been described previously (Baker et al., 1977). In the present study, exposure was from 1-4 weeks. The specimens ultimately used for analysis were developed in Kodak D 19. Cells associated with a minimum of 8 silver grains were photographed at random. The final print magnification was X 25,000.
Analysis of EM autoradiographs It is now well established that a majority of silver grains in EM autoradiographs do not directly overlie the point in the tissue section from which the isotopic decay particles responsible for their formation arose. This is the result of a number of factors, such as variable electron energy, section and emulsion thickness, angle of emission and above all, the ability of the electron microscope to resolve this 'range distribution' of silver grains. A number of approaches have been proposed which are designed to account for this 'cross-fire' from a source of decay which may be in organelle a, to the site of the silver grain which may be over organelle b. In the present study we have used the hypothetical grain methods of Blackett and Parry (1973, 1977). In this method, a computer programme is used to compare hypothetical grain distributions with the real grain distribution. According to the hypothesis being tested, hypothetical sources of radioactivity are established by applying to the autoradiographs a transparent overlay screen which contains computer-predicted source-to-site dis-
tances which generate the hypothetical silver grains 'emanating' from these sources. The directions of these hypothetical decays are, of course, random. The distances are derived from experimentally measured range distribution curves for line sources of a given isotope of the type first reported by Salpeter et al. (1969). Hence a rectangular matrix of cross-fire is created, there being invariably more sites than sources. Estimated of radioactivity in the various sources derive from systematic modification of the hypothetical source values by a minimising sub-routine until the hypothetical grain and real grain distribvtions fit optimally for the designated sites as assessed by the chi-square test. A feature of this method is that any hypothesis of radiolabel distribution can be tested. In cases where the crude location of silver grains is difficult to interpret by visual inspection, it is usual to collect hypothetical grain data by random or regular sampling so as to include all the potential sources of radioactivity in the analysis. In this study, however, many of the silver grains were sited close to the plasma membrane. Therefore, in order to collect sufficient relevant hypothetical grains for this source, which occupies a negligible area, regular points on the membrane were sampled for cross-fire (hypothetical grains) in addition to regular sampling of the rest of the cell for other potential sources. It should be stated that all of the real grains must be, and were, counted. This type of 'weighted' hypothetical grain sampling means that the relative area of each potential source is not realistically represented. Hence, only the computer estimates expressed as 'relative activity' (% of total, see Blackett and Parry, 1977) are valid. For the 4 experiments reported, the numbers of hypothetical and real grains counted ranged from 1161-1311 and 400-615 respectively.
Results
Retention of radioactivity by the EM preparative procedure The elution of counts from the radioiodinated cell suspension during fixation and dehydration was monitored by direct gamma counting of triplicate samples treated TABLE II HYPOTHETICAL GRAIN ANALYSIS OF LACTOPEROXIDASE A N D EXOGENOUS H202 Source
% Total radioactivity
Plasma membrane Cytosol Mitochondria Nucleus Matrix
54.7 + 4.7 17.3 + 2.8 1.8 ___1.1 7.7 + 1.4 18.5 ± 3.9
Total X2 1.9 (3 DF)
P > 0.5
125I-LABELLED
DAUDI
CELLS
USING
6 TABLE III HYPOTHETICAL GRAIN ANALYSIS OF 125I-LABELLED HUMAN LACTOPEROXIDASE AND EXOGENOUS HzOz Source
% Total radioactivity
Plasma membrane Cytosol Nucleus Matrix
67.0 _+4.5 16.2 + 5.5 4.3 _+2.5 12.6 + 1.1
Total X~ 1.3 (1 DF)
P > 0.20
PBLs USING
identically to those prepared for electron microscopy (Table I). A l t h o u g h 80% of radiolabel appeared to be lost i n the 2 fixation stages (glutaraldehyde; glutaraldeh y d e + o s m i u m tetroxide) this was unlikely to be a t t r i b u t a b l e to loss of p r o t e i n - b o u n d radioiodine. Rather, it was more likely to be the result of loss of radioactivity from the surface connected system of c o n t a m i n a t i n g platelets a n d radioiodine, possibly associated with lipid. The cacodylate rinses a n d ethanol d e h y d r a t i o n remove 5% of total radiolabel.
Fig. 1. EM autoradiograph of human peripheral blood lymphocytes incubated with Na~:51 in the presence of the exogenouslyadministered H202 reaction. Membrane disruption is obvious and may contribute to the grain value for the matrix. × 19,100. n = nucleus, m-mitochondria.
7 TABLE IV HYPOTHETICAL GRAIN ANALYSIS OF J251-LABELLED HUMAN PBLs USING G-GO TO GENERATE H202 Source
% Total radioactivity
Plasma membrane Cytosol Nucleus
92.5 +__1.8 2.7 _+ 1.0 4.9 ± 0.7
Total X2 = 5.3 (4 DF)
P > 0.20
(a) Subcellular location of radioiodide following lactoperoxidase/exogenous I-t,0, catalysed-labelling Daudi cells. T h e results o f the h y p o t h e t i c a l g r a i n analysis for the c u l t u r e d cell line are s h o w n in T a b l e II. O n l y 55% o f t h e t o t a l c e l l - a s s o c i a t e d r a d i o l a b e l was a t t r i b u t a b l e to the cell m e m b r a n e w i t h m o s t o f the r e m a i n d e r d i s t r i b u t e d e q u a l l y b e t w e e e n t h e c y t o s o l a n d ' m a t r i x ' . W h i l s t t h e n u c l e u s c o n t a i n e d a l m o s t 8% the m i t o c h o n d r i a a c c o u n t e d for less t h a n 2% o f the total. T h e e x c e l l e n c e o f t h e fit
Fig. 2. EM autoradiograph of human peripheral blood lymphocytes incubated with Na 1251 in the presence of the G-GO reaction. The cell and cell membrane appears well fixed and almost all the silver grains are associated with the cell membrane. × 20,000.
8
TABLE V HYPOTHETICAL GRAIN ANALYSIS OF 1251-LABELLED HUMAN PBLs USING G-GO TO GENERATE H 202 Source
% Total radioactivity
Plasma membrane Cytosol Nucleus
89.0 + 0.8 8.9 +_0.5 2.1 +_0.4
total X2 = 2.0 (2 DF)
P > 0.20
b e t w e e n real a n d final hypothetical grains ( P > 0.5) suggests that the ' m a t r i x ' is genuinely labelled. However, it seems p r o b a b l e that m u c h of this c o m p o n e n t represents i n c o m p l e t e l y sloughed-off m e m b r a n e a n d / o r 'glycocalyx' debris produced as a consequence of the labelling technique. Peripheral blood lymphocytes (PBLs). T a b l e III shows the results for the equivalent r a d i o i o d i n a t i o n a n d analysis of PBLs. I n this i n s t a n c e 67% of the radiolabel was located in the cell m e m b r a n e . The values for cytosol (16.2%), nucleus (4.3%) a n d matrix (12.6%) are of a similar m a g n i t u d e to those in the D a u d i cells (see T a b l e II).
Fig. 3. EM autoradiograph of human peripheral blood lymphocytes incubated with Na~25I alone. Although the cells are well fixed they are devoid of silver grains. × 10,000.
Once again, there is fine structural evidence to demonstrate that a certain level of cell membrane disruption may substantially contribute to the value for the matrix (Fig. 1).
(b) Subcellular location of radioiodide following lactoperoxidase/glucose-glucose oxidase labelling The estimates of subcellular radioactivity for PBLs radioiodinated with lactoperoxidase and the G-GO coupled reaction for generation of H202 are expressed for duplicate experiments (Tables IV and V). Since the initial computer results revealed that matrix and mitochondria were unlabelled, these sources were omitted from the repeat analyses shown in order to provide a model of subcellular labelling which would give the best fit as judged by the chi-square test. In both experiments approximately 90% of radioactivity was associated with the plasma membrane of the PBLs with the remainder in the cytosol or nucleus. The hypothetical grain analysis revealed that the matrix was not labelled. Moreover, the infrequency of fine structural membrane damage in these experiments (Fig. 2) supports our observation that only where such damage occurs will the matrix appear to be labelled.
(c) Control experiment When PBLs were incubated with Na125I alone, the resultant autoradiographs prepared in an identical manner to those described above were largely devoid of silver grains (Fig. 3). The few silver grains which were observed appeared to be non-specifically distributed.
Discussion
The study of membrane proteins of small defined cell populations requires radiolabelling of these macromolecules to facilitate analyses. A prerequisite of such radiolabelling techniques including lactoperoxidase-catalysed radioiodination is that exclusively plasma membrane proteins should be labelled. We have therefore investigated 2 radioiodination techniques to assess the fidelity with which this goal may be achieved. Previous EM autoradiographic studies on the distribution of radioiodine in cellular compartments of cells labelled by the lactoperoxidase method have shown that the grains are not exclusively located at the cell membrane in spite of claims made to the contrary (Marchalonis et al., 1971; Hubbard and Cohn, 1975). Indeed, measured range distribution data for 1251 (Fertuck and Salpeter, 1976; Blackett et al., 1980) show that depending upon the autoradiographic conditions and, in particular, the emulsion type, the spread of all the grains from a known linear source is several times the 'half-distance' value (Salpeter et al., 1969). A subsequent study (Gonatas et al., 1976) employed histograms of grain density, measured from the plasma membrane, to predict whether or not the image spread was consistent with labelling exclusively in the plasma membrane. They concluded that since the
l0 intracellular grain distribution extended beyond that predicted by the line source data of Fertuck and Salpeter (1976), a proportion of the radioactive decays (36%) must be intracellular in origin. It is, however, questionable whether the convoluted plasma membrane of lymphocytes can be legitimately considered as a linear source. Consequently, it was our objective to employ a method of analysis which could assess the effect of cross-fire from sources of irregular geometry. We believe that the use of the hypothetical grain method enables a more precise estimate of activity to be determined in all the significantly radiolabelled sources. The use of a albumin/glutaraldehyde gel (Bullock and Christian, 1976) as a support medium for EM processing of cell suspensions not only serves to eliminate the need for potentially damaging repeated centrifugation steps, but also provides a means of spacing the cells such that cross-fire from one to another can be easily minimised. Analysis of cells radiolabeUed by lactoperoxidase catalysed iodination, the hydrogen peroxide being supplied either exogenously or from the G-GO reaction, demonstrated that in both cases, the highest percentage of the radioisotope was localised in the plasma membrane. A greater proportion of plasma membrane-associated radioactivity was demonstrable in cells where the hydrogen peroxide was derived from the G - G O reaction (Tables IV and V) than in those cells where the radiolabelling reaction was promoted by exogenously added hydrogen peroxide (Tables II and lID; clearly less cytoplasmic radioactivity was evident in the former cell population than in the latter. We conclude from these data that the G-GO system for generation of hydrogen peroxide in the lactoperoxidase catalysed iodination provides a method of labelling plasma-membrane proteins of greater fidelity than the equivalent reaction driven by exogenously supplied hydrogen peroxide. A possible explanation for the above observations is that the relatively high concentration of hydrogen peroxide in systems employing exogenously added catalyst may alter the permeability of the plasma-membrane to small ions without adversely affecting cell viability in the short term. In coupled systems where hydrogen peroxide is generated biochemically, the latter is probably utilised immediately by lactoperoxidase, thus precluding generation of high local concentrations of a potentially membrane-oxidising agent. In other words, it does not seem valid to assume a close correlation between cell viability, as assessed by dye exclusion tests (Gonatas et al., 1976), and the degree of intracellular penetration of radioiodine.
References
Baker, J.R.J., G.R. Bullock,N. Crawford and D.G. Taylor, 1977, Am. J. Pathol., 88, 277. Blackett, N.M. and D.M. Parry, 1973, J. Cell Biol. 57, 9. Blackett, N.M. and D.M. Parry, 1977, J. Histochem. Cytochern.25, 206. Blackett, N.M., D.M, Parry and J.R.J. Baker, 1980, J. Histochem. Cytochem.28, 1050. BiSyum,A., 1968, Seand. J. Clin. Lab. Invest. 21 (Suppl. 97), 77. Bullock, G.R. and R.A. Christian, 1976, Histochem. J. 8, 291.
11 Fertuck, H.C. and M.M. Salpeter, 1976, J. Cell. Biol. 69, 144. Gonatas, N.K., J.O. Gonatas, A. Stieber, J.C. Antoine and S. Avrameas, 1976, J. Cell. Biol. 70, 477. Hubbard, A.L. and Z.A. Cohn, 1975, J. Cell. Biol. 64, 438. Marchalonis, J.J., 1969, Biochem. J. 113, 299. Marchalonis, J.J., R.E. Cone and V. Santer, 1971, Biochem. J. 124, 921. Mollenhauer, H.H., 1964, Stain Technol. 39, 111. Salpeter, M.M., L. Bachmann and E.E. Salpeter, 1969, J. Cell Biol. 41, 1.